Workflow pattern / Competitive intelligence

Build a content-intelligence workflow for editorial teams

Understand how a topic is being covered publicly, where the gaps are, and which formats recur.

Intermediate / Dataset / Weekly

The problem

Editorial planning runs on intuition about what already exists, which is expensive to check and quick to go stale.

Who it is for

Content and editorial teams planning technical or research-led publishing.

Teams: Editorial / Content strategy / Growth

Inputs

  • Topic and subtopic list
  • Source set
  • A content-attribute schema
  • Gap criteria

Conceptual process

  1. 01Collect

    Gather published pieces per subtopic from the defined source set.

  2. 02Attribute

    Extract format, depth, structure, and question addressed.

  3. 03Cluster

    Group by the question the piece answers rather than by keyword.

  4. 04Gap-find

    Identify questions with thin or no coverage.

  5. 05Brief

    Emit briefs that state the question, the gap, and the sources reviewed.

Flow diagram

  1. 01Topics
  2. 02Collect
  3. 03Attribute
  4. 04Cluster by question
  5. 05Find gaps
  6. 06Editorial brief

Example output

Illustrative output
{  "question": "How should retrieval handle version-specific documentation?",  "existing_coverage": 2,  "depth": "shallow",  "formats_seen": ["forum answer", "changelog note"],  "gap": "No structured guide found in reviewed sources",  "sources_reviewed": ["https://example.com/forum/1", "https://example.org/blog/2"]}

Data-quality considerations

  • Cluster by question, not by keyword — the same keyword serves different intents.
  • 'No coverage found in reviewed sources' is not the same as 'no coverage exists'.
  • Record the source set so the gap claim is bounded.

Failure modes

  • Gap identified because the source set was too narrow.
  • Format counted as depth — a long piece is not necessarily a thorough one.
  • Briefs that replicate existing coverage in a different order.

Suggested architecture

  • Topic registry
  • Collector
  • Attribute extractor
  • Question clustering
  • Gap analyser
  • Brief generator

What to test first

  1. 01Check the gap list against a subtopic you know well.
  2. 02Confirm each brief names its bounded source set.

Related reading